This application is in response to PA-98-052 for the Mentored Patient- Oriented Research Career Award (K23). The broad intersection of drugs and violence both directed toward oneself (suicide) and toward others (assault) remains a serious public health problem. A substantial proportion of these events are drug-related. The main objective of this application is to allow the candidate to develop a research career in the area of drug-related harm, specifically drug-related suicidal and/or homicidal behavior (hereafter, DRSHB). The mentor, Dr. Jack Mendelson, will provide a unique source of supervision and guidance for the candidate. The goals of the career development plan are to (1) develop the expertise necessary for the design and execution of clinical studies pertaining to DRSHB, (2) acquire further training in public health for research in clinical populations, (3) obtain the necessary qualities and skill for independence as a researcher. The primary aim of the research plan is to amplify our understanding of the impact of recent drug use (within 24 hours), co-morbid diagnosis (depression and alcohol), and drug combinations (cocaine and alcohol) on the course of DRSHB. Three prospective studies on psychiatric inpatients are specifically designed to rigorously examine DRSHB. Study One, Hypothesis: Suicide attempters with recent drug use will have more severe attempts and higher levels of suicidal/homicidal ideation than attempters without recent drug use. Study Two, Hypothesis I: Depressed alcoholics, compared to patients with major depression only and patients with alcohol dependence only, will have more severe attempts and higher levels of suicidal/homicidal ideation. Hypothesis II: For depressed alcoholics, the amount of recent alcohol use will be positively associated with the degree of suicidality/homicidality. Study Three, Hypothesis I: Patients with alcoholic and cocaine use disorders, compared to patients with alcohol use disorders only and cocaine use disorders only, will have more severe attempts and higher levels of suicidal/homicidal ideation. Hypothesis II: For patients with alcohol and cocaine use disorders, recent use of alcohol and cocaine will have more severe suicidal/homicidal behavior than either drug alone. The present studies as designed will extend and enhance our current knowledge of the phenomenology of DRSHB and shed light on the complex interplay of the drug-individual-environment triad. Understanding these relationships is essential for the development of useful risk-assessment instruments, appropriate intervention strategies and improved suicide and violence prevention.